391 
off from the middlemost part of the cells of the creeping filaments. Descending 
filaments are sometimes given off from the lower part of the erect filaments; they 
have the same appearance as the creeping ones and usually issue 
from the lower end of the cells while the erect branches are pro- 
duced at their upper end. Exceptions from this rule and transitional 
forms between the erect and the descending or creeping filaments 
may sometimes occur. Thus, creeping filaments may arise from the 
upper end of a cell, and erect filaments may change into a creep- 
ing one (fig. 330 A). 
The erect filaments are usually 10—15 w thick (6—17 w), and 
the thickness is essentially the same in all the Danish waters; at 
Bornholm, however, the thickness is on an average a little smaller, 
6—12 w. The cells contain a single nucleus and a number of small 
parietal chromatophores without pyrenoid (Kuckuck, |. c.). In the 
older transversal walls a refringent 
ring like that described by me in Rho- 
dochorton islandicum K. Rosenv.! may 
sometimes appear. 
A vegetative propagation is some- 
times realised by fragments of the 
thallus consisting of a branched fila- 
ment loosening from the mother plant 
and producing at the surface of frac- 
ture a downward growing filament 
(figs. 329, 330 E), becoming thus “‘cutt- 
ings” such as those described by me 
in Rhodochorton islandicum, |. c. p. 67. 
The middlemost cell in fig. 330 D is 
perhaps preparing the formation of 
Fig. 328. 
Rhodochorton Ro- 
thii, from Born- 
holm. Creeping 
filament giving off 
two erect filaments. 
The latter have 
been regenerated 
after the apical 
part has died. 
270 :1. 
such a cutting, giving rise to a small rhizoid penetrat- 
ing into the subjacent cell. These cuttings were 
principally found in specimens collected at Copen- 
hagen (Frederiksholms Kanal) in September.” 
The sporangia are usually clustered at the ends 
Fig. 329. 
Rhodochorton Rothü. Plant probably 
arisen by regeneration of an isolated part 
of the frond. 200 :1. 
! Note sur une Floridée aérienne. Bot. Tidsskr. Vol. 23, 1900, p. 67. 
of the erect filaments, as admirably represented in 
LE Jorıs Liste, pl. V, but they may also be more 
scattered in the upper end of these filaments. They 
? SvEDELIUS (Östersj. Algflora, 1901, p. 129) mentions the “cuttings” described by me in Rh. is- 
landieum as stolons; they are, however, no stolons, but fragments of ordinary filaments loosening by 
splitting of a transversal wall, and germinating by giving rise to creeping filaments. 
50 
